Multiple-Intent Text Separation Using Reinforcement Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current natural language processing (NLP) techniques fail to assign appropriate importance to context in customer-generated text, leading to biased insights, resource wastage, and incorrect recommendations due to overlapping contextual categories and temporal inconsistencies.
Innovation Solution
A context system using reinforcement learning to identify coreference and unrelated text, extract relevant phrases and sentiments, and generate final context text by processing with a reinforcement learning model to conserve resources and improve insights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current NLP techniques process all text data uniformly, then comprehensive text analysis is achieved, but computing resources are wasted and insights are biased due to overlapping contexts
Solution Approach 1:
The patent segments text data into distinct context categories (e.g., product review context vs. customer service context) and processes each segment with appropriate NLP techniques. This segmentation allows the system to assign different importance weights to different contexts, improving measurement precision while avoiding wasteful uniform processing of all text data
Solution Approach 2:
The patent applies local quality by treating different contexts with different processing qualities. Specifically, it assigns higher importance weights to relevant contexts (e.g., current product feedback in product reviews) and lower weights to less relevant contexts (e.g., old product feedback), thereby optimizing resource allocation and insight accuracy
2Reliability
If current NLP techniques analyze all contextual categories equally, then no context is missed, but incorrect recommendations result due to overlapping contexts and temporal inconsistencies
Solution Approach 1:
The patent performs preliminary action by pre-processing text data to identify and categorize different contexts before main analysis. It assigns importance weights to different contexts in advance, separating relevant from irrelevant information. This preliminary differentiation prevents context mixing and ensures that subsequent analysis produces reliable recommendations without losing essential context information
3Measurement precision
If current NLP techniques process text data without context separation, then processing is simple, but biased insights and incorrect recommendations occur
Solution Approach 1:
The patent segments text data into distinct context categories and processes each segment separately with appropriate NLP techniques. This segmentation improves insight accuracy by preventing context mixing while maintaining manageable system complexity through modular processing of separated contexts
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting importance weights assigned to different contexts based on their relevance. This allows the system to adapt processing focus to each specific use case (e.g., emphasizing current product feedback in product reviews vs. past customer service issues in support analysis), improving insight accuracy without requiring completely different systems for each scenario
Data Source
AI summary
A device identifies, in multi-context text data, unrelated text and coreference text, and extracts coreference clusters, coreference sentences, and coreference sentiments based on the coreference text. The device extracts unrelated sentences from the unrelated text, and assigns tenses to the coreference sentences and the unrelated sentences. The device extracts phrases and entities from the coreference sentences and unrelated sentences, and assigns tense flags that exclude present tense sentences. The device selects past tense phrases and future tense phrases, and combines the past tense phrases and the future tense phrases to generate phrases. The device identifies invalid phrases in the phrases, and identifies similarities between the coreference sentences and the invalid phrases. The device processes the coreference text, the coreference tenses, the coreference sentiments, and the similarities, with a reinforcement learning model, to generate final context text.